Doi10.1016 j.heliyon.2017.e00381

Metabolic and transcriptomic
analysis of Huntington’s
disease model reveal changes
in intracellular glucose levels
and related genes
Gepoliano Chaves 1
, Rıfat Emrah Özel 1
, Namrata V Rao, Hana Hadiprodjo,
Yvonne Da Costa, Zachary Tokuno, Nader Pourmand *
Biomolecular Engineering Department, University of California Santa Cruz, 1156 High Street, Santa Cruz, CA 95064,
USA
* Corresponding author.
E-mail address: Pourmand@soe.ucsc.edu (N. Pourmand).
1
G.C. and R.E.O. contributed equally.
Abstract
Huntington’s Disease (HD) is a neurodegenerative disorder caused by an
expansion in a CAG-tri-nucleotide repeat that introduces a poly-glutamine stretch
into the huntingtin protein (mHTT). Mutant huntingtin (mHTT) has been
associated with several phenotypes including mood disorders and depression.
Additionally, HD patients are known to be more susceptible to type II diabetes
mellitus (T2DM), and HD mice model develops diabetes. However, the
mechanism and pathways that link Huntington’s disease and diabetes have not
been well established. Understanding the underlying mechanisms can reveal
potential targets for drug development in HD. In this study, we investigated the
transcriptome of mHTT cell populations alongside intracellular glucose measure-
ments using a functionalized nanopipette. Several genes related to glucose uptake
and glucose homeostasis are affected. We observed changes in intracellular
glucose concentrations and identified altered transcript levels of certain genes
including Sorcs1, Hh-II and Vldlr. Our data suggest that these can be used as
markers for HD progression. Sorcs1 may not only have a role in glucose
Received:
7 February 2017
Revised:
2 July 2017
Accepted:
4 August 2017
Cite as: Gepoliano Chaves,
Rıfat Emrah Özel,
Namrata V Rao,
Hana Hadiprodjo,
Yvonne Da Costa,
Zachary Tokuno,
Nader Pourmand. Metabolic
and transcriptomic analysis of
Huntington’s disease model
reveal changes in intracellular
glucose levels and related
genes.
Heliyon 3 (2017) e00381.
doi: 10.1016/j.heliyon.2017.
e00381
http://dx.doi.org/10.1016/j.heliyon.2017.e00381
2405-8440/© 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).

metabolism and trafficking but also in glutamatergic pathways affecting trafficking
of synaptic components.
Keywords: Medicine, Neurology, Biochemistry, Cell biology, Neuroscience
1. Introduction
Huntington’s Disease (HD) is a progressive, autosomal dominant neurodegenera-
tive disease that produces physical, mental and emotional changes due to loss of
critically important brain neurons. The genetic basis for HD is an expansion of
cytosine-adenine-guanine (CAG) repeats in the huntingtin (HTT or IT15) gene that
leads to the formation of a prolonged polyglutamine (polyQ) tract in the N-terminal
region of the mutant huntingtin protein (mHTT). The wild type huntingtin protein
(HTT) is important for the intracellular transport and trafficking of proteins,
organelles and vesicles. The expansion of glutamine (>36 repeats) produces a gain
of function mHTT that affects the healthy function of cellular machinery,
ultimately resulting in neurotoxicity and detrimental cell lethality in the brain [1].
Considerable research efforts have been made to elucidate the molecular and
cellular mechanisms underlying HD pathology. The proposed mechanisms through
which mHTT causes neurodegeneration include mutant protein aggregation,
vesicle association, elevated oxidative stress, excitotoxicity, mitochondrial and
transcriptional dysregulation [2, 3]. Evidence for several of these mechanisms has
been found in post-symptomatic disease models or post-mortem brain samples.
Controversially, during the pre-symptomatic stage of HD, cellular architecture and
morphology have been found to be disrupted but neurodegeneration has been found
to be minimal or absent [4, 5]. These studies have been limited to
immunohistochemistry, fluorescence staining and immunoblot analysis, and
provided static information of HD [6, 7]. However, little is known at the early
stages of the disease about the dynamics of metabolic and transcriptomic changes,
which could be instrumental not only for HD therapy for diagnosed patients but
also for at-risk individuals. Striatum, a subcortical part of the forebrain, is the most
vulnerable part of the brain in HD, almost disappearing during the course of
disease. One of the major clinical symptoms of HD is the loss of the medium spiny
neurons in the striatum [8]. Therefore, although genome-wide gene expression
studies have been conducted on targeted affected tissues in the post mortem human
brain, it has proven difficult to detect specific changes in the human striatum [9].
To expand our understanding of changes in the transcriptional landscape of the
affected cells, mRNA expression profiling was performed on cells isolated from
the striatum of rat embryos expressing wild type and mutant HTT.
As shown by Lin and Beal, detecting changes in carbohydrates can be instrumental
for the development of useful therapies that slow down HD progression and reduce
its severity at the early stages of the disease [10]. Recent evidence suggests that,
Article No~e00381
2 http://dx.doi.org/10.1016/j.heliyon.2017.e00381

similar to other neurodegenerative diseases, mice exhibit increased deregulation in
brain energy metabolism at the late stages of HD [11]. A super-family of glucose
transporter genes, the GLUT gene family, is responsible for the transport and
uptake of glucose [12, 13]. The GLUT family includes 12 genes, each encoding
one GLUT protein. Various members of this gene family were detected in the
brain, although glucose transporter protein type I exclusively mediates transport
across the blood-brain barrier [14, 15]. GLUT1, initially thought to be the only
glucose transporter protein expressed in the brain, supports the basal metabolic
needs of proliferating cells [16]. GLUT2 is primarily GLUT protein expressed in
the pancreatic β-cells, liver and kidneys. In the pancreas, GLUT2 is believed to act
in conjunction with glucokinase in the glucose-sensing mechanism. GLUT4 is the
insulin-responsive receptor and is primarily expressed in the heart, adipose tissue,
and skeletal muscle, where it is responsible for the reduction in the glucose levels
in the post-prandial rise of glycaemia. Insulin stimulates the translocation of
intracellular vesicles containing GLUT4 to the plasma membrane, resulting in a
10- to 20-fold increase in glucose transport [17].
Current assessment methods for intracellular glucose levels in HD models are
limited and performed with a combination of techniques including gas
chromatography-mass spectrometry (GC-MS), liquid chromatography-mass spec-
trometry (LC-MS), uptake assays, and electrophysiology [18]. The invasiveness,
low sensitivity, and specificity of these techniques prevent longitudinal study of
biological model cells. Other major limitations of these techniques are the complex
sample preparation methods during which cellular concentration of glucose can be
altered. Therefore, accurate real time detection of glucose starting from the early
and going to the later stages of the disease may provide biochemical evidence that
can inform the design of novel drugs and therapeutic regimens. We recently
reported the use of nanopipette technology for intracellular measurements [19, 20].
Importantly, because nanosensors are minimally invasive and not destructive to the
cell, it is possible to take repeated measurements of the same cell. In this work, we
employed glucose nanosensors for intracellular glucose measurements in HD cell
models (ST14A) at different time points to investigate glucose metabolism over
time. In parallel, we conducted transcriptomic analysis to understand temporal
changes in RNA expression in model HD cells. We report impaired glucose
metabolism in single rat striatum mHTT cells using our nondestructive nano-
glucose sensor. This is the first report determining the intracellular glucose levels
in mHTT cells to be in the range of 0.8–1.5 mM. A 2.5 fold decrease in
intracellular glucose levels in HD cells indicated an altered glucose metabolism
that was further investigated by transcriptomic study. Significant differences were
seen in glucose uptake and glucose homeostasis genes. Interestingly, there was an
increased expression of Sorcs1 in mHTT cells which has been previously
associated with late onset of Alzheimer’s disease, another neurodegenerative
Article No~e00381

disorder. Overexpression of Sorcs1 was confirmed by performing RT-qPCR which
corroborates Sorcs1 can be used as a biomarker for HD progression.
2. Materials and methods
2.1. Reagents
Glucose Oxidase (GOx) from Aspergillus niger ≥100,000 units/g (Type VII,
lyophilized, EC Number 232-601-0), ferrocene (98%), poly-l-lysine (PLL) (0.1%
solution in water), glutaraldehyde (GA) (25% in water) and D-glucose were
purchased from Sigma Aldrich (St. Louis, MO). Silver wires (125 μm) were
supplied from A-M Systems (Sequim, WA). Glucose free Dulbecco’s Modified
Eagle Medium (DMEM), trypsin (0.25%, phenol red) and penicillinstreptomycin
were bought from Gibco while Hank’s Balanced Salt Solutions and fetal bovine
serum were purchased from GE Healthcare (GE Healthcare). 500 μm gridded
plates were obtained from Ibidi (Ibidi USA, Inc., Madison, WI).
2.2. Glucose nanosensor fabrication
Glucose nanosensors with an outside diameter of 1.00 mm and an inside diameter
of 0.70 mm were fabricated from quartz capillaries (Sutter Instrument, Novato,
CA) using a P-2000 laser puller (Sutter Instrument, Novato, CA) as described
elsewhere [19]. Briefly, to create glucose nanosensors, the nanopipettes were
modified as follows: Pulled nanopipettes were backfilled with a 15 μl solution
containing 10 mM Ferrocene prepared in 100 mM PBS (pH 7, supplemented with
0.1 M KCl). An Ag/AgCl electrode was placed into the nanopipette as a working
electrode while another Ag/AgCl electrode was immersed in the cell media as a
reference/counter electrode. Nanopipettes were modified with PLL by backfilling
the nanopipette interior. Then nanopipette surface was treated with a 10% (v/v)
solution of glutaraldehyde (GA) for 30 min. GOx (900 mU) was then reacted with
the activated nanopipette walls. All glucose nanosensors were calibrated in DMEM
with standard glucose concentrations.
2.3. Cell culture
The Huntington’s disease model cell lines that were used in this work are both
striatal cell lines derived from rat embryos and obtained from Coriell Institute cell
repository. The two cell lines express human HTT fragments of 15 and 120
polyglutamine repeats representing wild type (ST14A-Q15) and mHTT disease
model (ST14A-Q120), respectively. All cells were cultivated and maintained in
DMEM (supplemented with 10% fetal bovine serum, antibiotics) with various
glucose concentrations. DMEM was supplemented with 1 nM insulin when needed
to study effects of insulin.
Article No~e00381

2.4. Intracellular glucose measurement
The intracellular glucose measurement setup consists of an inverted microscope
Olympus IX 70 with Spot Insight CMOS camera to image cells. The nanosensors
are fixed to a microscope by a pipette holder (Axon Instruments). The holder is
connected to an Axopatch 700B amplifier (Molecular Devices) for current
measurement, an MP-285 micromanipulator (Sutter Instrument) for coarse control
of the nanopipette positioning in the X, Y, and Z directions, a Nanocube piezo
actuator (Physik Instrument) for fine control in the X, Y, and Z directions, and a
PCIe-7851R Field Programmable Gate Array (FPGA) (National Instruments) for
hardware control of the system. The system is operated using custom-coded
software written in LabVIEW. Current-clamp technique has been used at 1nA with
signal filter at 1 kHz. The signal was further digitized by an Axon Instruments
Digidata 1322A. The data were then recorded by a LabVIEW 9.0 home-made
software, as described previously [21]. While in the medium, a fixed potential of
500 mV was applied to the glucose nanosensor. This biased potential generates an
ion current through the liquid-liquid interface that is used as the input into a
feedback loop analyzed by the scanning ionic conductance microscope. When a
5–10% current decrease is detected, the glucose sensor is penetrated into the target
cell up to 0.8 μm at 100mm/s. The sensor is maintained inside the target cell for a
pre-defined time of 60 seconds and then withdrawn to the initial position. Each
condition was tested on a minimum of 9 cells. Three consecutive readings were
recorded for each individual target cell to ensure reproducibility and robustness. In
order to investigate underlying molecular mechanism of these observed changes, a
detailed transcriptome study was performed.
2.5. RNA isolation from HD cells
RNA extraction and purification were performed using RNeasy Mini kit from
Qiagen. Prior to RNA isolation, HD cells were collected with conventional trypsin
treatment and counted using Bio-Rad automated cell counter. A minimum of 5 ×
105
cells were used for RNA isolation. Samples were stabilized by treatment with
RNAlater (Ambion). After RNA isolation, nucleic acid quantity and quality were
checked using NanoDrop (Thermo Fisher Scientific).
2.6. RNA-seq library preparation, qRT-PCR and sequencing
2.6.1. cDNA synthesis
Extracted total RNA was converted to cDNA using the Smart-seq2 protocol and
the cDNA was preamplified as described by a previoud study [22]. Briefly, the
cDNA product was mixed with KAPA HiFi HotStart ReadyMix (2X, KAPA
Biosystems), ISPCR primer (10uM) and nuclease-free water (Gibco) and PCR
amplified as follows: 98 °C 3 min, then 12 cycles of (98 °C 15 s, 67°C 20 s, 72 °C
Article No~e00381

6 min), with a final extension at 72 °C for 5 min. The cDNA was cleaned up using
a 1X ratio of AMPure XP beads (Beckman Coulter).
2.6.2. Quantitative real-time PCR
Real time PCR experiments were performed using 500 ng of reverse transcribed
RNA and analyzed with SYBR green qPCR master mix (kapa Biosystems) using
an Mx3000P instrument (Stratagene). The relative quantification of Sorcs1
expression levels was performed using a previously described method [23]. For
these experiments actb was used as a reference gene. Sorcs1 and actb specific
primer sequences designed in previous studies were used (Table 1).
2.6.3. Library preparation and sequencing
The cDNA was diluted to ∼300 pg/ul and then processed with the Nextera XT
DNA library preparation kit according to the manufacturer’s protocol (Illumina).
These libraries were purified and size-selected with an average library size of
300–600 bp as determined by Bioanalyzer 2100 high-sensitivity DNA assay
(Agilent).
Functional library concentration was determined with the KAPA Biosystems
library quantification kit. The libraries were denatured after quantification and
loaded on Illumina HiSeq 2000 for sequencing.
2.6.4. Quality control and mapping of sequencing reads
Sequencing adapter sequences were removed from reads using Cutadapt version
1.4.2 [24]. Quality of preprocessed reads was evaluated using FastQC (http://www.
bioinformatics.babraham.ac.uk/projects/fastqc/). The preprocessed reads were
mapped as paired-end reads using the STAR pipeline [25] with default parameters
against the UCSC rn5 Rattus norvegicus genome, combined with the human
huntingtin gene (http://hgdownload.soe.ucsc.edu/goldenPath/rn5/bigZips/).
2.6.5. Gene expression analysis
HTSeq package was used to generate the gene matrix from the output sam file
produced by the STAR alignment. A gtf file containing Ensembl identifiers as well
Table 1. Sequences of primers used in this study.
GENE FORWARD REVERSE REFERENCE
Sorcs1 5'-AGCCAACAGAAATAAACCTTTCC-3’ 5'-TAATGTGGTCTTCTTCTTGATGT −3' [76]
β-Actin 5’-CACCCGCGAGTACAACCTTC-3’ 5’-CCCATACCCACCATCACACC-3’ [77]
Article No~e00381

as gene names was used for rat and human genes. For the ERCC genes, the gtf file
from Thermo Fisher Scientific was downloaded from the company website (www.
thermofisher.com). The gene matrix was then subjected to DESeq2 differential
expression analysis, as described [26, 27, 28, 29]. Benjamini-Hochberg multiple
testing adjustment was used as a method to deal with p-values that are low due to
chance, and not because the measurement represents a significant value. False
Discovery Rate values used were calculated by DESeq2 package. Hierarchical
Clustering of differentially expressed genes was performed using the Pretty
Heatmaps function in R (v. 3.2.3). Pearson coefficients were used as correlation
coefficients for the Hierarchical Cluster Analysis heatmaps. The Panther
classification system (www.pantherdb.org) was used to visualize the pathways in
which differential genes were involved.
2.7. Statistical analysis
Analysis of Variance (ANOVA) and independent t-test were performed using
Prism (GraphPad) software when indicated.
3. Results and discussion
The energy metabolism in presymptomatic and symptomatic HD individuals is
known to be defective [30, 31]. Reduced glucose uptake in the striatum and cortex
of HD patients prior to the onset of clinical symptoms has been previously revealed
by positron emission tomography (PET) scanning. Additionally, this hypometa-
bolism demonstrated a high correlation with the onset and progression of HD,
independent of the extent of CAG expansion [32]. Recently, Besson et al. reported
that the increase of glucose transporters was beneficial to HD pathology in a
Drosophila melanogaster model [33]. However, due to the destructive nature of
previously employed intracellular glucose measurement methods, a direct
interrogation of temporal changes at the transcriptomic and metabolic level could
not have been studied. In this work, we took a step forward by deploying
functionalized nanopipettes as glucose sensors to measure changes in intracellular
glucose concentrations over time, and performed RNA sequencing to evaluate
concomitant changes in expression of genes related to glucose metabolism.
3.1. Glucose nanosensors reveal temporal intracellular glucose
changes at HD cells
Direct monitoring of glucose in cells is difficult due to the small size of the
biological specimen and the complexity of the intracellular environment. Glucose
nanosensors are made of glucose oxidase (GOx) functionalized quartz nanopipette
with ∼ 100 nm pores. Due to their small size, glucose nanosensors can be inserted
into individual cells, providing a non-destructive and direct in vitro quantification
Article No~e00381

tool. We have recently demonstrated the use of this nanosensor in conjunction with
a customized cell finder at normal and cancer cells [19].
Typical physiological concentrations of glucose in biological systems are in the
low-micromolar range [34]; intracellular glucose levels in HD cells is unknown.
More importantly, to our knowledge, there are no reports on the assessment of
temporal changes in intracellular glucose concentration in HD models, an
assessment that could illuminate one aspect of the cellular pathophysiology of
HD. To achieve the goal of measuring intracellular glucose over time, we placed
nanosensors in the cytoplasm of wild type (WT) (ST14A-Q15) and HD (ST14A-
Q120) cells.
Cells underwent consecutive passages in fresh culture medium and glucose
measurements were made at the end (day 3 (T1), day 6 (T2), day 11 (T3), day 16
(T4)) of each passage. In the initial experimental conditions, WT and HD cells
were cultured in low glucose media (LGM) and high glucose media (HGM)
containing 0.5X and 1X (4.5 g/L) glucose respectively. Cells were also cultured in
glucose-free media but neither WT nor HD cells survived in this condition.
Therefore, measurements could not be conducted in the absence of glucose. It is
important to note that the regular culture condition of these cells does not require or
include insulin.
Exposure of WT cells to low and high glucose media in the absence of insulin
resulted in similar intracellular glucose levels as measured by glucose nanosensors
(Figs. 1 and 2 ). Intracellular glucose levels were found to be almost stable from
[(Fig._1)TD$FIG]
Fig. 1. Bar graphs showing the average intracellular glucose concentrations measured using GOx-
functionalized nanosensor in DMEM with low (LGM) glucose content in the absence of insulin (INS
(−)) and 1 nM insulin (INS (+)) in wild type (WT) and disease cell lines (DM). Nine to 12 individual
cells with 3 replicates were tested for each condition. The measurements were performed at day 3 (T1),
day 6 (T2), day 11(T3), day 16 (T4).
Article No~e00381

T2 to T4, and the range was between 1.79 to and 2.58 mM. The most significant
difference was observed for the T1 time point where intracellular levels were about
3 fold lower than those of T2-T4 in wild type cells. This low concentration can be
due to either fast metabolism or rapid growth rate of young cells. When the glucose
nanosensor was used on HD cells, we observed that intracellular levels of glucose
were about 2.5 fold lower than that found WT cells for T3 and T4 (Figs. 1 and 2).
To compare glucose uptake in WT and HD cells, we exposed the cells to 1 nM
insulin (normal non-fasting insulin level), as insulin regulates glucose uptake by
activating glucose transporters 1 and 4 (GLUT1 and GLUT4) [35]. We saw a
general increase in intracellular glucose levels for all time points for both cell lines
(Figs. 1 and 2). As expected, in HGM, intracellular glucose levels were much
higher than in LGM. Notably, the effect of insulin was more pronounced at T3 and
T4 suggesting that glucose uptake of HD cells is facilitated when insulin is present.
The average intracellular glucose concentrations and their variations for both cell
types under all conditions are summarized in Table 2.
3.2. Assessing transcriptomics in rat HD cells
We integrated transcriptomics and metabolomics i.e., the metabolome provided
phenotypic (glucose) measurements to which we anchored the global measure-
ments of the transcriptome related to glucose pathways. In order to achieve this,
RNA-seq transcriptomics was performed to complement the data obtained from
glucose nanosensor measurements, using a similar set-up. RNA was extracted from
[(Fig._2)TD$FIG]
Fig. 2. Bar graphs displaying the average intracellular glucose concentrations measured using GOx-
functionalized nanosensor in DMEM with high glucose (HGM) content in the absence insulin (INS(−))
and 1 nM insulin (INS (+)) for wild type (WT) and disease cell lines (DM). Nine to 12 individual cells
with 3 replicates were tested for each condition. The measurements were performed at day 3 (T1), day 6
(T2), day 11 (T3), day 16 (T4).
Article No~e00381

Table 2. Summary of intracellular glucose measurements at wild type (WT) and disease (HD) cells in the absence (INS −) and presence (INS +) of insulin.
The standard deviations (σ) were calculated for n = 9 replicate measurements for each condition (σ = (1/N(Σ(χi− μ)2
)1/2
).
WT HD WT HD WT HD WT HD
T1 T2 T3 T4
[Glucose] mM LGM INS (-) 0.41 (±0.06) 0.92 (±0.25) 1.79 (±0.43) 1.45 (±0.26) 2.53 (±0.13) 0.93 (±0.13) 2.54 (±0.29) 0.84 (±0.29)
LGM INS (+) 1.17 (±0.05) 1.34 (±0.35) 2.15 (±0.55) 1.58 (±0.31) 2.61 (±0.16) 1.89 (±0.33) 2.35 (±0.29) 1.36 (±0.29)
HGM INS (-) 0.72 (±0.16) 0.77 (±0.23) 2.44 (±0.87) 1.28 (±0.23) 2.58 (±0.23) 1.42 (±0.16) 2.49 (±0.23) 0.11 (±0.51)
HGM INS (+) 1.52 (±0.36) 1.29 (±0.29) 2.67 (±0.16) 1.80 (±0.31) 3.12 (±0.40) 1.66 (±0.40) 3.19 (±0.18) 2.16 (±0.17)
ArticleNo~e00381
10http://dx.doi.org/10.1016/j.heliyon.2017.e00381
2405-8440/©2017TheAuthors.PublishedbyElsevierLtd.ThisisanopenaccessarticleundertheCCBYlicense

wild type (WT) (ST14A-Q15) and disease (HD) (ST14A-Q120) cells. The gene
expression levels were assessed across four passages T1-T4.
First, to assess the sequencing by synthesis (SBS) method as a quantitative
profiling tool, ERCC spike-in controls were included in the cDNA mix with the
samples. ERCC spike-in controls showed Pearson correlation above 90% in all
library-to-library comparisons. Also, a 90% Pearson correlation was observed in
the expression of ERCC pseudo-genes to the ERCC starting concentrations
provided by the ERCC manufacturer (Fig. 3 (A)). Although Spearman correlation
was slightly lower (minimum 88%) (Fig. 3 (B)), Pearson correlation is a better
parameter for our dataset because it assumes a linear relationship between ERCC
gene expression measurements and ERCC starting concentration [36].
Comparisons of upregulated, downregulated and significantly different genes were
made with the top 50 genes in a universe of 3213 differentially expressed genes. A
recent RNAseq study in human post-mortem HD patients also showed 5480 genes
differentially expressed, indicating a similar order of magnitude in the number of
differentially expressed genes in rat and humans due to mHTT [9].
3.3. Glucose-related genes were differentially expressed in
mHTT cells
Specific genes used for the analysis were involved in glycolysis, trichloroacetic
acid cycle (TCA), pentose phosphate pathway (PPP) and glucose transport;
[(Fig._3)TD$FIG]
Fig. 3. A) Pearson correlation of the starting ERCC spike-in concentration and the ERCC spike-in
count after library preparation and RNA-seq. ERCC expression values were calculated as log2 of
counts. B) Spearman correlation of the starting ERCC spike-in concentration and the ERCC spike-in
counts after RNA-seq. ERCC expression values were calculated as log2 of counts.
Article No~e00381

accessed from a list of genes in glucose metabolism provided in Qiagen’s website
(www.qiagen.com). The pattern of expression of genes involved in these pathways
in mutant and wildtype were plotted as a heatmap using the Pheatmap package
available in R [75]. Seven genes in the glycolysis pathway, eleven genes in the
TCA, two genes in the PPP, and one gene expressing a glucose transporter were
found to be affected by mHTT in this study (Fig. 4).
3.4. Survival effect of HK-II and Glut1 on HD cells
Hexokinase catalyzes the first step of glycolysis and represents a regulatory
enzyme [74]. Expression of hexokinase II (HK-II) was found to be upregulated in
our rat striatal HD cells, a finding also observed by other authors [37]. An
increased expression of hexokinase II has also been associated with other
neurodegenerative disorders [38] (Fig. 4). It was demonstrated previously that
alteration in HK-II gene expression has a direct impact on glucose metabolism
[39].
Hexokinases (HKs) phosphorylate the glucose transported through glucose
transporters (GLUTs) on the plasma membrane to produce glucose-6-phosphate
(G-6P) (Berg et al., 2002). HK activity is inhibited by G-6P providing a feedback
[(Fig._4)TD$FIG]
Fig. 4. Glucose metabolism related genes derived from glycolysis (blue), PPP (red), TCA (green) and
glucose transport pathways (purple) differentially expressed in wild type and HD cells. Statistical
significance was determined by the Benjamini-Hochberg multiple testing adjustment, as described in
DESeq2 documentation, based on p-value comparison. FDR was applied on p-values for significance
cut-off.
Article No~e00381

mechanism [40]. The reduced glucose content measured in mutant cells could be
attributed to the increased expression of HK-II in HD cells. HK-II upregulation has
been previously shown to be an important consequence of metabolic re-
programming in cancer [39].
The expression of Glut1 (Slc2a1), the main class of glucose receptor involved in
glucose transport from the blood-brain barrier to the CNS [15], was also
upregulated in the mutant cell type (Fig. 4). A previous study in a hematopoietic
cell line showed that increased glucose phosphorylation due to co-expression of
HK-II and Glut1 has an anti-apoptotic effect. The protective effect was attributed
to increase in NADPH activity through the pentose phosphate pathway eliciting an
anti-apoptotic effect [41]. The mutant huntingtin protein causes neuronal
dysfunction and eventual cell death. Some of the pathways found to be abnormal
in Huntington’s disease models are transcriptional impairment, neuronal
excitotoxicity, oxidative damage, inflammation, apoptosis, and mitochondrial
dysfunction. The anti-apoptic effect could slow down the disease progression in
our HD rat cell model, similar to observations by the study of Johri and Beal [42].
In summary, our data suggest a compensation mechanism by which increased
expression of HK-II and Glut1, may drive mutant cells to survive huntingtin
mutation.
Insulin treatment is known to increase HK-II mRNA and protein in various cell
types and the increase is blocked by inhibition of PI3K, an upstream kinase of Akt,
as well as inhibition of mTORC1 [39]. mTORC1 is a protein that integrates signals
to regulate cell growth and metabolism.
Genes coding for IDH1, ENO3, MDH2 and ALDOC proteins were also found to
be differentially expressed (Fig. 4) in mHTT cells. Eno3 is a gene that codes the
enolase version of the muscle. The deficiency of this protein correlates with
glycogen storage disease XIII [43]. Together, the results of this glucose
homeostasis targeted analysis suggest that glucose metabolism is altered in our
HD model, in agreement with other studies [44].
3.5. Upregulated and downregulated genes in HD rat cell model
To further understand the perturbation initiated by huntingtin, we analyzed the
most abundantly differentiated genes of our cell model in the Panther Classification
System (www.pantherdb.org). The group of the top 50 upregulated genes was
segmented after the pathway enrichment analysis of genes present using Panther
Classification. Fig. 5 shows the heatmap and pie-chart analysis of the up-regulated
genes. Fig. 5b shows the main cellular pathways identified by the Panther
Classification System among the top 50 expressed genes. They were identified to
be cellular receptors controlling signal transduction, including EGF receptor, Wnt
signaling, and glutamate receptor. Pathways related to Huntington’s and
Article No~e00381

Alzheimer’s diseases were also found in the Panther Classification. Igf1 is
involved in two pathways: the MAPK pathway, related to cell development and
differentiation, and the AKT/PKB pathway that controls insulin metabolic actions
(Fig. 5) [45].
SORCS1 protein is necessary for insulin granule secretion in β-cells [46], however
was not found annotated with such function by the Panther Classification System
in our analysis. SORCS1 is sortilin-related vacuolar protein sorting 10 (VPS-10)
domain containing receptor 1. Interestingly, the gene expressing SORCS1 controls
[(Fig._5)TD$FIG]
Fig. 5. a) Top 50 up-regulated genes in the comparison log2 (value in mutant)/(value in wild-type), as
described in the DESeq2 documentation. b) Pie chart showing Panther pathway analysis of the top 50
up-regulated genes of rat cells expressing human huntingtin.
Article No~e00381

protein trafficking and transport of other cellular components, such as the amyloid
beta plaques [47, 48], and has been established as an important hallmark of
Alzheimer’s disease. Gene Sorcs1 was found upregulated in the mutant type in this
study (Fig. 5a). A recent study suggested that SORCS1 has a novel regulatory
mechanism and acts as a modulator of sortillin function [49]. Sortillin, also a
member of the Vps 10 protein sorting receptor family, is involved in biological
processes such as glucose and lipid metabolism. It has been proposed that
dysfunction of SORCS1 protein may contribute to both the APP/Aβ disturbance
underlying Alzheimer’s disease (AD) and the insulin/glucose disturbance in
Diabetes Mellitus (DM) [50].
To better investigate changes observed in mHTT model in this study, we compared
transcriptomics data from other Huntington mice models and included some of the
changes observed in AD. Out of 70 genes annotated in the Alzheimer’s disease
presinilin pathway, 26 genes were found altered in mHTT cells (Fig. 5 and Fig. 6).
The presinilin pathway is dysregulated in AD, cleaving several single-transmem-
brane proteins within the membrane domain, including the amyloid precursor
protein, Notch, ErbB4, E-cadherin, Nectin-1alpha, and CD44 (www.pantherdb.
org). Notably, another important AD marker, gene Apoe, was affected in our cell
model. Interestingly, Vldlr, an Apoe receptor was the top downregulated gene in
our study (Fig. 7). As mentioned before, HK-II, a glycolysis enzyme, was affected
by the mHTT protein. Besides the already established low-glucose metabolism
characteristic of AD, it was observed that AD condition affects the regulation of 6-
phosphofructo-2-kinase in human patients, further illustrating common features
between AD and HD [51].
Other pathways significantly affected in our mHTT model, including calcium
signaling, glutamate receptor activity and synaptic transmission, were observed by
other researchers in HD mice and human patients [52, 53, 54, 55]. For instance,
Achour and collaborators identified Gata2 transcription factor, a gene involved in
establishment of tissue-specific enhancers, which was found to be affected in the
present study, highly expressed in the striatum, and not cerebellum, of HD mice
model [56]. Furthermore, genetic variations in intron 1 of Sorcs1 have been
associated with Alzheimer’s disease in several studies [57, 58]. In order to confirm
upregulation of Sorcs1 observed in RNA seq analysis, we stimulated cells with
insulin as described in intracellular glucose measurements and performed gene
specific PCR analysis (Fig. 8). Sorcs1 was found to be upregulated in mHTT
compared to wild type as observed in RNA-seq data. To understand the effect of
insulin and/or glucose on expression of Sorcs1 in mHTT cells, qRT-PCR was
performed. The mHTT cells not treated with glucose and insulin showed the
highest Sorcs1 expression (Fig. 9). Significant difference was observed when these
mHTT cells (no insulin, no glucose) were compared to WT cells grown in all
conditions tested. However, we observed no statistically significant difference in
Article No~e00381

Sorcs1 expression in mutant cells treated with either insulin or glucose. Our results
suggest that mHTT mutation was the only factor causing higher Sorcs1 expression.
However, more stringent and detailed analysis will enable better understanding of
the effect of additional glucose and insulin on expression of Sorcs1.
In response to insulin, sortillins can transport GLUT4 to the membrane for uptake
of glucose [59]. There are no previous studies demonstrating that the two proteins,
SORCS1 and sortilin, can act identically. However, recent literature has indicated
that the SORCS are sortilin-related CNS expressed proteins, and, as they belong to
the same subgroup of VSP10 receptor family, these proteins are related [46, 47,
60]. Since SORCS1 and sortilins belong to the same family of proteins, we were
[(Fig._6)TD$FIG]
Fig. 6. Number of differentially expressed genes per signaling pathway according to Pantherdb
analysis. A minimum of 10 genes per signaling pathway was requested. Pathways were organized from
highest to lowest number of differentially expressed genes within each pathway. Some of the pathways
in red bars represent signaling pathways which are discussed in this manuscript.
Article No~e00381

interested in understanding whether increased Sorcs1 in cells with Huntington
phenotype would affect Glut4 gene expression, thereby increasing glucose uptake.
As Glut4 expression was not significantly different in HD and normal cells in the
RNA-Seq analysis, we performed protein expression studies. Our results showed
that GLUT4 protein was expressed at lower level in rat HD cells (Fig. 10A and B).
Basal levels of GLUT4 along with other glucose receptors including GLUT1, the
main class of glucose receptor, may compensate for the lack of GLUT4 protein
observed in mHTT cells. Nevertheless, other factors including post-transcriptional
[(Fig._7)TD$FIG]
Fig. 7. a) Top 50 down-regulated genes in the comparison log2 (value in mutant)/(value in wild-type),
as described in the DESeq2 documentation. b) Pie chart showing Panther pathway analysis of the top 50
down-regulated genes of rat cells expressing human huntingtin.
Article No~e00381

modification or action of miRNA may also have an effect on decreased GLUT4
protein levels, which needs to be investigated further. Hou and Pessin reported that
overexpression of sortillin stabilizes GLUT4 protein, increases the formation of
insulin responsive compartments, and promotes insulin-stimulated glucose uptake
[(Fig._8)TD$FIG]
Fig. 8. Polymerase Chain Reaction products of SorCS1 and actin primers amplification. 5μL of actin
amplicons and 5 μL of SorCS1 amplicon were mixed and allowed to resolve in agarose gel
electrophoresis as indicated above. Insulin and glucose treatments were indicated above each lane.
Results suggest a higher expression of SorCS1 in the mHTT cell type, confirming RNA-seq data. The
full, uncropped gel is available as Supplementary Fig. 6.
[(Fig._9)TD$FIG]
Fig. 9. Relative gene expression (fold change) of Sorcs1 gene in mHTT and wild type was normalized
to reference condition (wild type cells grown in media containing no additional glucose or insulin). actB
was used as endogenous control. mHTT cells had 2- 4-fold increase in Sorcs1 levels. Results are
presented as mean (T1, T2, T3 and T4) ± SE. Statistical analysis was assessed by unpaired t-test and
one-way ANOVA. **P < 0.05 was considered statistically significant.
Article No~e00381

[59]. Higher GLUT4 and/or higher metabolism could explain the increased
intracellular glucose concentration in wild-type cells and a time-dependent study
could give a better idea whether GLUT4 is indeed regulated by SORCS1 in
addition to sortillin, as observed in the study by Hou and Pessin [59]. As GLUT4
levels were lower in mHTT cells, we became interested in analyzing GLUT1
levels, the main class of glucose receptor. Results showed higher GLUT1 levels in
mHTT cells suggesting that GLUT1 could indeed compensate for the lack of Glut4
in mHTT cells (Fig. 10C).
The fine regulation of brain developed neutrophic factor (BDNF) by sortillin has
been implicated in neuronal and tumor cell survival. Interestingly, BDNF protein
was also shown as a diabetes and Huntington’s disease marker [2, 61]. In our cell
model, following the trends described by Zuccato and collaborators, Bdnf levels
were lower in mutant cells than in wild type. Furthermore, it has been
demonstrated that BDNF protein is one of the molecules that is a target of
organelle sorting mediated by SORCS1 [62]. BDNF, once transported to the
striatum from the cortex, induces cholesterol synthesis and is related to synaptic
plasticity and neuronal survival [63]. Sortillin has also been implicated in LDL-
cholesterol metabolism and VLDL secretion [64, 65]. The very low density
lipoprotein receptor (Vldlr) was found as the top downregulated gene in HD cells
as compared to wild type cells in our data set (Fig. 7). VLDLR is an apolipoprotein
[(Fig._10)TD$FIG]
Fig. 10. Western blot of glucose receptors. Glucose transporter 4 protein expression comparing mHTT
and wild type HTT-expressing cell lines corresponding to passages 1 (T1) (A) and 4 (T4), (B) of ST14A
cells. A decreased GLUT4 expression is observed due to mHTT. C) Glucose transporter 1 protein
expression comparing mHTT and wild type HTT-expressing cell lines corresponding to passages 1 (T1)
(C) and 2 (T2), (D) of ST14A cells. Mutants (mHTT cells) had an increased expression of GLUT1
receptor. The full, uncropped blots are available as Supplementary Figs. 8A1, 8A2, 8B1, 8B2, 8C1,
8C2, 8D1, 8D2.
Article No~e00381

E (APOE) receptor and the interaction of these two proteins has important
implications in lipid metabolism [66]. Downregulation of Vldlr gene indicates that
lipid metabolism is also impaired in mutant huntingtin striatal cells, as has been
shown by another group [67]. Other genes that were found to be downregulated in
this study are shown in Fig. 7a, along with the pathways they belonged to in
Fig. 7b.
Together, these data suggest the involvement of Sorcs1, insulin, Glut4 and Glut1 in
the HD condition, at least in our cell model, providing molecular candidates that
support a decreased intracellular glucose concentration due to mutant huntingtin.
To our knowledge, no other group has reported the involvement of Sorcs1, an
Alzheimer’s disease and diabetes marker in a Huntington’s disease model. Sorcs1
may not only be involved in impairment of glucose metabolism but also lipid
metabolism in Huntington’s disease.
3.6. Other pathways affected by mutant huntingtin
Using transcriptomics and metagenomics, we sought to understand how mHTT
influences not only HD related pathways but also other major pathways. In order to
understand the pathways affected, we classified differentially expressed genes
according to their signaling pathways, with at least 10 genes observed per pathway
in the Pantherdb classification system. 1145 genes (out of 3213) are represented in
Fig. 6. The signaling pathway most significantly affected by mutated huntingtin in
number of genes was the wingless-related (Wnt) signaling pathway (P00057),
showing 65 disregulated genes. Beta-catenins, a canonical element from the Wnt
pathway, forms a complex with GSK-3B; they regulate epidermal growth factor
and insulin, as well as control cell fate. Other genes that were differentially
expressed were important in the insulin/MAPK/PKB/IGF pathways (P00032 and
P00033). 64 genes were impaired in the gonadotrophin-releasing hormone receptor
(P06664). Dysregulation of inflammation mediated by the cytokines and
chemokines pathway (P00031) supports the anti-inflammatory treatments sug-
gested for human HD [9]. Glutamatergic pathways were also found to be affected
in our study (P00037 and P00039). Glutamate ionotropic receptor AMPA type
subunit 3 (gria 3), a gene in this pathway, was downregulated in our study (Fig. 7).
Morton et al. also showed this gene to be downregulated in R2/6 mice, affecting
synaptic plasticity [68]. A recent study showed SORCS1 regulating the AMPA
receptors [47]. All the other pathways affected due to mutant huntingtin
(containing 10 or more than 10 genes per pathway) are shown in Fig. 6. In
summary, data shown on Fig. 6 support general observations from the literature
about Huntington’s disease, as well as provide support to our observation of Sorcs1
being dysregulated due to huntingtin mutation, in that this gene interacts with
glutamate receptors [47].
Article No~e00381

4. Conclusion
Correlation between HD and glucose blood levels has been extensively studied.
Although some studies do not show a correlation between HD and glucose
metabolism, [3], other studies suggest the involvement of insulin resistance
mediated by the huntingtin mutation and cross-talk between glucose metabolism
and other metabolic pathways. Among these pathways, involvement of fatty-acid
biosynthesis and cholesterol synthesis are described in HD progression [63, 67, 69,
70, 71, 72, 73]. Here, we showed impaired glucose metabolism in rat striatum cells
using a functionalized nanopipette as a non-destructive intra-cellular glucose
measurement technique. Impaired glucose metabolism along with altered
expression of genes concerned with energy metabolism contributes to HD
pathogenesis. A new additional target, SorCS1 was identified to be altered in HD
cells along with previously reported transporter related genes, glut1 and glut4.
Increased levels of SorCS1 correlate positively with loss of the ability of the cell to
sort proteins like glut4 properly. Further work on generating a knock-out mutant of
SorCS1 will shed more light on better understanding the role of this gene in
Huntington’s disease. Glut1 and HK-II expression may have a beneficial effect on
HD pathology by delaying apoptosis. The induction of genes in the PPP could
possibly delay HD progression, thereby providing efficient neuroprotection against
reactive oxygen species. These data represent a comprehensive integration of
metabolic and transcriptomic expression data in rat HD cells. This research yields
important advances in our understanding of HD pathogenesis. It remains to be
shown whether the changes observed here are similar across different cell types
within the brain of affected human individuals.
Declarations
Author contribution statement
Gepoliano Chaves, Rifat Emrah Ozel: Conceived and designed the experiments;
Performed the experiments; Analyzed and interpreted the data; Wrote the paper.
Namrata Rao V: Performed the experiments; Analyzed and interpreted the data;
Wrote the paper.
Hana Hadiprodjo: Analyzed and interpreted the data.
Yvonne Da Costa, Zachary Tokuno: Performed the experiments.
Nader Pourmand: Conceived and designed the experiments; Analyzed and
interpreted the data; Contributed reagents, materials, analysis tools or data; Wrote
the paper.
Article No~e00381

Funding statement
This work was supported in part by grants from the National Institutes of Health
[P01-35HG000205], National Institute of Neurological Disorders and Stroke
[R21NS082927], Prize money awarded through the NIH Follow That Cell
Challenge, internal support from UCSC Jack Baskin Engineering Dean and
CAPES, Coordination for the Improvement of Higher Education Personnel -
Brazil.
Competing interest statement
The authors declare no conflict of interest.
Additional information
Supplementary content related to this article has been published online at http://dx.
doi.org/10.1016/j.heliyon.2017.e00381.
Acknowledgements
We acknowledge Mr. Akshar Lohith for technical support, and Professor Doug
Kellogg's laboratory for use of the Western blot equipment.
References
[1] S.E. Andrew, Y.P. Goldberg, B. Kremer, H. Telenius, J. Theilmann, S. Adam,
et al., The relationship between trinucleotide (CAG) repeat length and clinical
features of Huntington’s disease, Nat. Genet. 4 (1993) 398–403.
[2] C. Zuccato, A. Ciammola, D. Rigamonti, B.R. Leavitt, D. Goffredo, L. Conti,
et al., Loss of huntingtin-mediated BDNF gene transcription in Huntington’s
disease, Science 293 (2001) 493–498.
[3] R. Nambron, E. Silajdžić, E. Kalliolia, C. Ottolenghi, P. Hindmarsh, N.R.
Hill, et al., A Metabolic Study of Huntington’s Disease, PLoS One 11 (2016)
e0146480.
[4] H. Mizuno, H. Shibayama, F. Tanaka, M. Doyu, G. Sobue, H. Iwata, et al.,
An autopsy case with clinically and molecular genetically diagnosed
Huntington’s disease with only minimal non-specific neuropathological
findings, Clin. Neuropathol. 19 (2000) 94–103.
[5] E. Gómez-Tortosa, M.E. MacDonald, J.C. Friend, S.A. Taylor, L.J. Weiler, L.
A. Cupples, et al., Quantitative neuropathological changes in presymptomatic
Huntington’s disease, Ann. Neurol. 49 (2001) 29–34.
Article No~e00381

[6] T. Pancani, D.J. Foster, M.S. Moehle, T.J. Bichell, E. Bradley, T.M. Bridges,
et al., Allosteric activation of M4 muscarinic receptors improve behavioral
and physiological alterations in early symptomatic YAC128 mice, Proc. Natl.
Acad. Sci. USA 112 (2015) 14078–14083.
[7] J. Modregger, N.A. DiProspero, V. Charles, D.A. Tagle, M. Plomann,
PACSIN 1 interacts with huntingtin and is absent from synaptic varicosities
in presymptomatic Huntington’s disease brains, Hum. Mol. Genet. 11 (2002)
2547–2558.
[8] X. Liang, J. Wu, P. Egorova, I. Bezprozvanny, An automated and quantitative
method to evaluate progression of striatal pathology in Huntington’s disease
transgenic mice, J. Huntingtons Dis. 3 (2014) 343–350.
[9] A. Labadorf, A.G. Hoss, V. Lagomarsino, J.C. Latourelle, T.C. Hadzi, J.
Bregu, et al., RNA Sequence Analysis of Human Huntington Disease Brain
Reveals an Extensive Increase in Inflammatory and Developmental Gene
Expression, PLoS One 10 (2015) e0143563.
[10] M.T. Lin, M.F. Beal, Mitochondrial dysfunction and oxidative stress in
neurodegenerative diseases, Nature 443 (2006) 787–795.
[11] A.I.I. Acuña, M. Esparza, C. Kramm, F.A. Beltrán, A.V. Parra, C. Cepeda,
et al., A failure in energy metabolism and antioxidant uptake precede
symptoms of Huntington’s disease in mice, Nat. Commun. 4 (2013) 2917.
[12] W.I. Sivitz, S.L. DeSautel, T. Kayano, G.I. Bell, J.E. Pessin, Regulation of
glucose transporter messenger RNA levels in rat adipose tissue by insulin,
Mol. Endocrinol. 4 (1990) 583–588.
[13] E.M. Quistgaard, C. Löw, F. Guettou, P. Nordlund, Understanding transport
by the major facilitator superfamily (MFS): structures pave the way, Nat.
Rev. Mol. Cell Biol. 17 (2016) 123–132.
[14] H. Lund-Andersen, Transport of glucose from blood to brain, Physiol. Rev.
59 (1979) 305–352.
[15] W.M. Pardridge, R.J. Boado, C.R. Farrell, Brain-type glucose transporter
(GLUT-1) is selectively localized to the blood-brain barrier. Studies with
quantitative western blotting and in situ hybridization, J. Biol. Chem. 265
(1990) 18035–18040.
[16] S.J. Vannucci, F. Maher, I.A. Simpson, Glucose transporter proteins in brain:
delivery of glucose to neurons and glia, Glia 21 (1997) 2–21.
[17] I.S. Wood, P. Trayhurn, Glucose transporters (GLUT and SGLT): expanded
families of sugar transport proteins, Br. J. Nutr. 89 (2003) 3–9.
Article No~e00381

[18] R. Kaddurah-Daouk, K.R.R. Krishnan, Metabolomics: a global biochemical
approach to the study of central nervous system diseases, Neuropsychophar-
macology 34 (2009) 173–186.
[19] R.A. Nascimento, R.E.E. Özel, W.H. Mak, M. Mulato, B. Singaram, N.
Pourmand, Single Cell Glucose Nanosensor Verifies Elevated Glucose Levels
in Individual Cancer Cells, Nano Lett. 16 (2016) 1194–1200.
[20] R.E.E. Özel, A. Lohith, W.H. Mak, N. Pourmand, Single-cell intracellular
nano-pH probes, RSC Adv. 5 (2015) 52436–52443.
[21] R. Adam Seger, P. Actis, C. Penfold, M. Maalouf, B. Vilozny, N. Pourmand,
Voltage controlled nano-injection system for single-cell surgery, Nanoscale 4
(2012) 5843–5846.
[22] S. Picelli, O.R. Faridani, A.K. Björklund, G. Winberg, S. Sagasser, R.
Sandberg, Full-length RNA-seq from single cells using Smart-seq2, Nat
Protoc. 9 (2014) 171–181.
[23] M.W. Pfaffl, A new mathematical model for relative quantification in real-
time RT-PCR, Nucleic Acids Res. 29 (2001) e45.
[24] C. Chen, S.S. Khaleel, H. Huang, C.H. Wu, Software for pre-processing
Illumina next-generation sequencing short read sequences, Source Code Biol.
Med. 9 (2014) 8.
[25] A. Dobin, C.A. Davis, F. Schlesinger, J. Drenkow, C. Zaleski, S. Jha, et al.,
STAR: ultrafast universal RNA-seq aligner, Bioinformatics 29 (2013) 15–21.
[26] M.I. Love, S. Anders, V. Kim, W. Huber, RNA-Seq workflow: gene-level
exploratory analysis and differential expression, F1000Res. 4 (2015) 1070.
[27] M.I. Love, W. Huber, S. Anders, Moderated estimation of fold change and
dispersion for RNA-seq data with DESeq2, Genome Biol. 15 (2014) 550.
[28] S. Anders, W. Huber, Differential expression analysis for sequence count
data, Genome Biol. 11 (2010) R106.
[29] W. Huber, V.J. Carey, R. Gentleman, S. Anders, M. Carlson, B.S. Carvalho,
et al., Orchestrating high-throughput genomic analysis with Bioconductor,
Nat. Methods 12 (2015) 115–121.
[30] F. Mochel, P. Charles, F. Seguin, J. Barritault, C. Coussieu, L. Perin, et al.,
Early energy deficit in Huntington disease: identification of a plasma
biomarker traceable during disease progression, PLoS One 2 (2007) e647.
[31] V.B. Mattis, C.N. Svendsen, Modeling Huntington's disease with patient-
derived neurons, Brain Res. (2015).
Article No~e00381

[32] A. Ciarmiello, G. Giovacchini, S. Orobello, L. Bruselli, F. Elifani, F.
Squitieri, 18F-FDG PET uptake in the pre-Huntington disease caudate affects
the time-to-onset independently of CAG expansion size, Eur. J. Nucl. Med.
Mol. Imaging 39 (2012) 1030–1036.
[33] M.T. Besson, K. Alegría, P. Garrido-Gerter, L.F. Barros, J.-C.C. Liévens,
Enhanced neuronal glucose transporter expression reveals metabolic choice in
a HD Drosophila model, PLoS One 10 (2015) e0118765.
[34] P. Mergenthaler, U. Lindauer, G.A. Dienel, A. Meisel, Sugar for the brain:
the role of glucose in physiological and pathological brain function, Trends
Neurosci. 36 (2013) 587–597.
[35] P. Ebeling, H.A. Koistinen, V.A. Koivisto, Insulin-independent glucose
transport regulates insulin sensitivity, FEBS Lett. 436 (1998) 301–303.
[36] M. Scott, D. Flaherty, J. Currall, Statistics: are we related? J. Small Anim.
Pract. 54 (2013) 124–128.
[37] A.J. Cooper, K.F. Sheu, J.R. Burke, W.J. Strittmatter, J.P. Blass,
Glyceraldehyde 3-phosphate dehydrogenase abnormality in metabolically
stressed Huntington disease fibroblasts, Dev. Neurosci. 20 (1998) 462–468.
[38] J.C. Corona, A. Gimenez-Cassina, F. Lim, J. Díaz-Nido, Hexokinase II gene
transfer protects against neurodegeneration in the rotenone and MPTP mouse
models of Parkinson’s disease, J. Neurosci. Res. 88 (2010) 1943–1950.
[39] D.J. Roberts, S. Miyamoto, Hexokinase II integrates energy metabolism and
cellular protection: Akting on mitochondria and TORCing to autophagy, Cell
Death Differ. 22 (2015) 364.
[40] P.H. Ratz, K.M. Berg, N.H. Urban, A.S. Miner, Regulation of smooth muscle
calcium sensitivity: KCl as a calcium-sensitizing stimulus, Am. J. Physiol.
Cell Physiol. 288 (2005) C769–83.
[41] J.C. Rathmell, C.J. Fox, D.R. Plas, P.S. Hammerman, R.M. Cinalli, C.B.
Thompson, Akt-directed glucose metabolism can prevent Bax conformation
change and promote growth factor-independent survival, Mol. Cell. Biol. 23
(2003) 7315–7328.
[42] A. Johri, M.F. Beal, Antioxidants in Huntington’s disease, Biochim. Biophys.
Acta 1822 (2012) 664–674.
[43] M. Vorgerd, J. Zange, Carbohydrate oxidation disorders of skeletal muscle,
Curr. Opin. Clin. Nutr. Metab. Care 5 (2002) 611–617.
[44] M. Ribeiro, T. Rosenstock, A. Oliveira, C. Oliveira, C. Rego, Insulin and
IGF-1 improve mitochondrial function in a PI-3K/Akt-dependent manner and
Article No~e00381

reduce mitochondrial generation of reactive oxygen species in Huntington’s
disease knock-in striatal cells, Free Radic. Biol. Med. (2014) 129–144.
[45] C.M. Taniguchi, B. Emanuelli, C.R. Kahn, Critical nodes in signalling
pathways: insights into insulin action, Nat. Rev. Mol. Cell Biol. 7 (2006)
85–96.
[46] M.A. Kebede, A.T. Oler, T. Gregg, A.J. Balloon, A. Johnson, K. Mitok, et al.,
SORCS1 is necessary for normal insulin secretory granule biogenesis in
metabolically stressed β cells, J. Clin. Invest. 124 (2014) 4240–4256.
[47] J.N. Savas, L.F.F. Ribeiro, K.D. Wierda, R. Wright, L.A. DeNardo-Wilke, H.
C. Rice, et al., The Sorting Receptor SorCS1 Regulates Trafficking of
Neurexin and AMPA Receptors, Neuron 87 (2015) 764–780.
[48] R.F. Lane, P. St George-Hyslop, B.L. Hempstead, S.A. Small, S.M.
Strittmatter, S. Gandy, Vps10 family proteins and the retromer complex in
aging-related neurodegeneration and diabetes, J. Neurosci. 32 (2012)
14080–14086.
[49] J.V. Larsen, G. Hermey, E.S. Sørensen, T. Prabakaran, E.I. Christensen, J.
Gliemann, et al., Human sorCS1 binds sortilin and hampers its cellular
functions, Biochem. J. 457 (2014) 277–288.
[50] R.F. Lane, J.W. Steele, D. Cai, M.E. Ehrlich, A.D. Attie, S. Gandy, Protein
sorting motifs in the cytoplasmic tail of SorCS1 control generation of
Alzheimer’s amyloid-β peptide, J. Neurosci. 33 (2013) 7099–7107.
[51] S. Sekar, J. McDonald, L. Cuyugan, J. Aldrich, A. Kurdoglu, J. Adkins, et al.,
Alzheimer’s disease is associated with altered expression of genes involved in
immune response and mitochondrial processes in astrocytes, Neurobiol.
Aging 36 (2015) 583–591.
[52] T. Seredenina, R. Luthi-Carter, What have we learned from gene expression
profiles in Huntington’s disease? Neurobiol. Dis. 45 (2012) 83–98.
[53] C. Tourette, B. Li, R. Bell, S. O’Hare, L.S. Kaltenbach, S.D. Mooney, et al.,
A large scale Huntingtin protein interaction network implicates Rho GTPase
signaling pathways in Huntington disease, J. Biol. Chem. 289 (2014)
6709–6726.
[54] C. Cepeda, M.A. Ariano, C.R. Calvert, J. Flores-Hernández, S.H. Chandler,
B.R. Leavitt, et al., NMDA receptor function in mouse models of Huntington
disease, J. Neurosci. Res. 66 (2001) 525–539.
Article No~e00381

[55] R. Luthi-Carter, A. Strand, N.L. Peters, S.M. Solano, Z.R. Hollingsworth, A.
S. Menon, et al., Decreased expression of striatal signaling genes in a mouse
model of Huntington’s disease, Hum. Mol. Genet. 9 (2000) 1259–1271.
[56] M. Achour, S. Le Gras, C. Keime, F. Parmentier, F.-X.X. Lejeune, A.-L.L.
Boutillier, et al., Neuronal identity genes regulated by super-enhancers are
preferentially down-regulated in the striatum of Huntington’s disease mice,
Hum. Mol. Genet. 24 (2015) 3481–3496.
[57] H.-F.F. Wang, J.-T.T. Yu, W. Zhang, W. Wang, Q.-Y.Y. Liu, X.-Y.Y. Ma,
et al., SORCS1 and APOE polymorphisms interact to confer risk for late-
onset Alzheimer’s disease in a Northern Han Chinese population, Brain Res.
1448 (2012) 111–116.
[58] C. Reitz, S. Tokuhiro, L.N. Clark, C. Conrad, J.-P.P. Vonsattel, L.-N.N.
Hazrati, et al., SORCS1 alters amyloid precursor protein processing and
variants may increase Alzheimer’s disease risk, Ann. Neurol. 69 (2011)
47–64.
[59] J.C. Hou, J.E. Pessin, Ins (endocytosis) and outs (exocytosis) of GLUT4
trafficking, Curr. Opin. Cell Biol. 19 (2007) 466–473.
[60] A. Nykjaer, T.E. Willnow, Sortilin: a receptor to regulate neuronal viability
and function, Trends Neurosci. 35 (2012) 261–270.
[61] B. Boyuk, S. Degirmencioglu, H. Atalay, S. Guzel, A. Acar, A. Celebi, et al.,
Relationship between levels of brain-derived neurotrophic factor and
metabolic parameters in patients with type 2 diabetes mellitus, J. Diabetes
Res. 2014 (2014) 978143.
[62] G.R. Lewin, B.D. Carter, Neurotrophic factors. Preface, Handb. Exp.
Pharmacol. 220 (2014) v–vi.
[63] V. Leoni, C. Caccia, The impairment of cholesterol metabolism in
Huntington disease, Biochim. Biophys. Acta 1851 (2015) 1095–1105.
[64] A. Strong, Q. Ding, A.C. Edmondson, J.S. Millar, K.V. Sachs, X. Li, et al.,
Hepatic sortilin regulates both apolipoprotein B secretion and LDL
catabolism, J. Clin. Invest. 122 (2012) 2807–2816.
[65] A. Strong, D.J. Rader, Sortilin as a regulator of lipoprotein metabolism, Curr.
Atheroscler. Rep. 14 (2012) 211–218.
[66] D.M. Holtzman, J. Herz, G. Bu, Apolipoprotein E and apolipoprotein E
receptors: normal biology and roles in Alzheimer disease, Cold Spring Harb.
Perspect. Med. 2 (2012) a006312.
Article No~e00381

[67] M. Valenza, D. Rigamonti, D. Goffredo, C. Zuccato, S. Fenu, L. Jamot, et al.,
Dysfunction of the cholesterol biosynthetic pathway in Huntington’s disease,
J. Neurosci. 25 (2005) 9932–9939.
[68] A.J. Morton, M.J. Hunt, A.K. Hodges, P.D. Lewis, A.J. Redfern, S.B.
Dunnett, et al., A combination drug therapy improves cognition and reverses
gene expression changes in a mouse model of Huntington’s disease, Eur. J.
Neurosci. 21 (2005) 855–870.
[69] L.A. Farrer, Diabetes mellitus in Huntington disease, Clin. Genet. 27 (1985)
62–67.
[70] K.A. Tobin, H.H. Steineger, S. Alberti, O. Spydevold, J. Auwerx, J.A.
Gustafsson, et al., Cross-talk between fatty acid and cholesterol metabolism
mediated by liver X receptor-alpha, Mol. Endocrinol. 14 (2000) 741–752.
[71] M. Valenza, E. Cattaneo, Cholesterol dysfunction in neurodegenerative
diseases: is Huntington’s disease in the list? Prog. Neurobiol. 80 (2006)
165–176.
[72] M.T. Nakamura, B.E. Yudell, J.J. Loor, Regulation of energy metabolism by
long-chain fatty acids, Prog. Lipid Res. 53 (2014) 124–144.
[73] E.J. Stephenson, J.A. Hawley, Mitochondrial function in metabolic health: a
genetic and environmental tug of war, Biochim. Biophys. Acta. 1840 (2014)
1285–1294.
[74] J.M. Berg, J.L. Tymoczko, L. Stryer, Biochemistry, fifth ed., W. H. Freeman,
New York, 2002.
[75] R Core Team, R: A language and environment for statistical computing, R
Foundation for Statistical Computing, Vienna, Austria, 2013. http://www.R-
project.org/.
[76] J. Lazar, et al., SORCS1 contributes to the development of renal disease in
rats and humans, Physiol. Genomics 45 (16) (2013) 720–728.
[77] S. Khowal, et al., Assessment of the therapeutic potential of hesperidin and
proteomic resolution of diabetes-mediated neuronal fluctuations expediting
Alzheimer's disease, RSC Adv. 5 (58) (2015) 46965–46980.
Article No~e00381

Doi10.1016 j.heliyon.2017.e00381

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Similar to Doi10.1016 j.heliyon.2017.e00381

Similar to Doi10.1016 j.heliyon.2017.e00381 (20)

Recently uploaded

Recently uploaded (20)

Doi10.1016 j.heliyon.2017.e00381